SimEpiNEM {epiNEM} | R Documentation |
Compares different network reconstruction algorithm on simulated data.
SimEpiNEM(runs = 10, do = c("n", "e"), random = list(FPrate = 0.1, FNrate = c(0.1, 0.5), single = 3, double = 1, reporters = 10, replicates = 2), maxTime = FALSE, forcelogic = TRUE, epinemsearch = "greedy", bnemsearch = "genetic", ...)
runs |
number simulation runs |
do |
string vector of algorithms to compare: e (epiNEM), n (Nested Effects Models), b (B-NEM), p (PC algorithm), a (Aracne), e.g. c("e", "n", "p") |
random |
list of false poitive rate FPrate, false negative rates FNrate, number of single knock-downs single, number of double knock-downs double, number of effect reporters reporters and number of replicates replicates |
maxTime |
TRUE if the algorithms are bound to a maximum running time in respect to epiNEM |
forcelogic |
if TRUE the randomly sampled ground truth network includes a complex logic with probability 1 |
epinemsearch |
greedy or exhaustive search for epiNEM |
bnemsearch |
genetic or greedy search for B-NEM |
... |
additional parameters |
returns list of specificity and sensitivity of inferred edges (spec, sens) and inferred expected data (spec2, sens2) and accuracy of logics (logics) and running time (time)
Martin Pirkl
res <- SimEpiNEM(runs = 1)